README.md

metadata

license: mit
tags:
  - generated_from_trainer
datasets:
  - cord
model-index:
  - name: cord-repo
    results: []

cord-repo

This model is a fine-tuned version of SCUT-DLVCLab/lilt-roberta-en-base on the cord dataset. It achieves the following results on the evaluation set:

• Loss: 0.2317
• Menu.cnt: {'precision': 0.9521739130434783, 'recall': 0.9733333333333334, 'f1': 0.9626373626373628, 'number': 225}
• Menu.discountprice: {'precision': 0.6, 'recall': 0.6, 'f1': 0.6, 'number': 10}
• Menu.nm: {'precision': 0.9011406844106464, 'recall': 0.9404761904761905, 'f1': 0.920388349514563, 'number': 252}
• Menu.num: {'precision': 1.0, 'recall': 0.8181818181818182, 'f1': 0.9, 'number': 11}
• Menu.price: {'precision': 0.9565217391304348, 'recall': 0.9758064516129032, 'f1': 0.9660678642714571, 'number': 248}
• Menu.sub Cnt: {'precision': 0.875, 'recall': 0.8235294117647058, 'f1': 0.8484848484848485, 'number': 17}
• Menu.sub Nm: {'precision': 0.6666666666666666, 'recall': 0.8125, 'f1': 0.7323943661971831, 'number': 32}
• Menu.sub Price: {'precision': 1.0, 'recall': 0.7, 'f1': 0.8235294117647058, 'number': 20}
• Menu.unitprice: {'precision': 0.9253731343283582, 'recall': 0.9117647058823529, 'f1': 0.9185185185185185, 'number': 68}
• Sub Total.discount Price: {'precision': 0.8571428571428571, 'recall': 0.8571428571428571, 'f1': 0.8571428571428571, 'number': 7}
• Sub Total.etc: {'precision': 1.0, 'recall': 0.75, 'f1': 0.8571428571428571, 'number': 8}
• Sub Total.service Price: {'precision': 1.0, 'recall': 1.0, 'f1': 1.0, 'number': 12}
• Sub Total.subtotal Price: {'precision': 0.8205128205128205, 'recall': 0.927536231884058, 'f1': 0.870748299319728, 'number': 69}
• Sub Total.tax Price: {'precision': 1.0, 'recall': 0.9555555555555556, 'f1': 0.9772727272727273, 'number': 45}
• Total.cashprice: {'precision': 0.9393939393939394, 'recall': 0.8732394366197183, 'f1': 0.9051094890510948, 'number': 71}
• Total.changeprice: {'precision': 0.9661016949152542, 'recall': 0.95, 'f1': 0.957983193277311, 'number': 60}
• Total.creditcardprice: {'precision': 0.8235294117647058, 'recall': 0.875, 'f1': 0.8484848484848485, 'number': 16}
• Total.emoneyprice: {'precision': 0.5, 'recall': 0.5, 'f1': 0.5, 'number': 2}
• Total.menuqty Cnt: {'precision': 0.71875, 'recall': 0.7666666666666667, 'f1': 0.7419354838709677, 'number': 30}
• Total.menutype Cnt: {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 8}
• Total.total Etc: {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 4}
• Total.total Price: {'precision': 0.9019607843137255, 'recall': 0.9292929292929293, 'f1': 0.9154228855721392, 'number': 99}
• Overall Precision: 0.9125
• Overall Recall: 0.9201
• Overall F1: 0.9163
• Overall Accuracy: 0.9355

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 300
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Menu.cnt	Menu.discountprice	Menu.nm	Menu.num	Menu.price	Menu.sub Cnt	Menu.sub Nm	Menu.sub Price	Menu.unitprice	Sub Total.discount Price	Sub Total.etc	Sub Total.service Price	Sub Total.subtotal Price	Sub Total.tax Price	Total.cashprice	Total.changeprice	Total.creditcardprice	Total.emoneyprice	Total.menuqty Cnt	Total.menutype Cnt	Total.total Etc	Total.total Price	Overall Precision	Overall Recall	Overall F1	Overall Accuracy
0.6711	2.0	200	0.2317	{'precision': 0.9521739130434783, 'recall': 0.9733333333333334, 'f1': 0.9626373626373628, 'number': 225}	{'precision': 0.6, 'recall': 0.6, 'f1': 0.6, 'number': 10}	{'precision': 0.9011406844106464, 'recall': 0.9404761904761905, 'f1': 0.920388349514563, 'number': 252}	{'precision': 1.0, 'recall': 0.8181818181818182, 'f1': 0.9, 'number': 11}	{'precision': 0.9565217391304348, 'recall': 0.9758064516129032, 'f1': 0.9660678642714571, 'number': 248}	{'precision': 0.875, 'recall': 0.8235294117647058, 'f1': 0.8484848484848485, 'number': 17}	{'precision': 0.6666666666666666, 'recall': 0.8125, 'f1': 0.7323943661971831, 'number': 32}	{'precision': 1.0, 'recall': 0.7, 'f1': 0.8235294117647058, 'number': 20}	{'precision': 0.9253731343283582, 'recall': 0.9117647058823529, 'f1': 0.9185185185185185, 'number': 68}	{'precision': 0.8571428571428571, 'recall': 0.8571428571428571, 'f1': 0.8571428571428571, 'number': 7}	{'precision': 1.0, 'recall': 0.75, 'f1': 0.8571428571428571, 'number': 8}	{'precision': 1.0, 'recall': 1.0, 'f1': 1.0, 'number': 12}	{'precision': 0.8205128205128205, 'recall': 0.927536231884058, 'f1': 0.870748299319728, 'number': 69}	{'precision': 1.0, 'recall': 0.9555555555555556, 'f1': 0.9772727272727273, 'number': 45}	{'precision': 0.9393939393939394, 'recall': 0.8732394366197183, 'f1': 0.9051094890510948, 'number': 71}	{'precision': 0.9661016949152542, 'recall': 0.95, 'f1': 0.957983193277311, 'number': 60}	{'precision': 0.8235294117647058, 'recall': 0.875, 'f1': 0.8484848484848485, 'number': 16}	{'precision': 0.5, 'recall': 0.5, 'f1': 0.5, 'number': 2}	{'precision': 0.71875, 'recall': 0.7666666666666667, 'f1': 0.7419354838709677, 'number': 30}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 8}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 4}	{'precision': 0.9019607843137255, 'recall': 0.9292929292929293, 'f1': 0.9154228855721392, 'number': 99}	0.9125	0.9201	0.9163	0.9355

Framework versions

• Transformers 4.29.2
• Pytorch 1.13.1
• Datasets 2.12.0
• Tokenizers 0.13.2